4 ms·
Many Americans already don't understand progressive tax as a concept. How can we realistically expect them to understand ranked choice voting?
by Freedom2 3y ago
Many Americans already don't understand progressive tax as a concept. How can we realistically expect them to understand ranked choice voting?
- tacon 3y agoProof by example: Alaska already does ranked choice voting https://www.elections.alaska.gov/RCV.php https://www.elections.alaska.gov/RCV.php
- gnicholas 3y agoTrue, but that doesn't actually establish that voters understand how it works. IIRC, they've only had it for a short time, so there's not much of a track record to establish that voters are casting votes in ways that are consistent with their beliefs and desires.
- edgyquant 3y ago[flagged]
- ACow_Adonis 3y agoThe same way other nations do?
- foofie 3y agoIt's vital that a democratic nation fully understands the method they use to elect representatives. People need to understand how their vote maps to a preference, and how their votes are tallied to determine who wins and who doesn't. If a significant share of the electorate does not understand the vote tallying process and expects an outcome different from the one that actually resulted from counting the votes, you will get accusations of vote fraud and election tampering.
- flir 3y agoSeems like you've already got that problem with FPTP. (What I'm getting at is that the problem's not the voting system, it's something else).
- psd1 3y agoMaybe, but on a maslow pyramid of healthy democracy, that would be a tier above an informed electorate and critical thinking skills. If you figure out how to develop those, please include Britain on your speaking tour.
- healthdare 3y ago[dead]
- godelski 3y ago1) Americans don't understand progressive tax because they are constantly being bombarded with information as to what it means. If you watch Fox news daily you will have a misunderstanding on the matter. Take care to make generalizations when you see things like knowledge differences divided between party lines. People aren't different, but the information they get is. 2) Many social choice experts advocate for cardinal systems over ordinal. Approval is quite popular due to its simplicity and effectiveness while Star or Score improve on alignment but all of these are substantially better than the current system. Cardinal has the advantage in the tabulation being that you do not need multiple elections (exception of Star, which is strictly 2 elections), and you can do a parallelized reduce sum tabulation. So algorithmically it is easy to understand (sum columns/candidates, pick largest total: argmax(column sum)). That has implications for transparency and election security. Some people are convinced these methods are harder for people to understand but I'd refer you back to point 1 and add that we're also very used to these systems as we're currently using a cardinal voting system on HN.
- psd1 3y agoIt is plain from the stories of fox news orphans that information diet can ruin good people, but I refuse to believe that people aren't different. I rebut your thesis thus: Oskar Schindler.
- godelski 3y agoWhat do you think the main difference in people is? Biological? Because I'm saying environmental. I'm not sure why you reference Schindler. Certainly there is a distribution and I'm not claiming people are clones (I don't think you are either). So outliers aren't proof of anything, to either argument.
- psd1 3y agoI think Dawkins' model of memes is helpful. I'm going to use freighted language and examples for brevity. Readers, please be generous; I'm doing conservatism because it's the topic on hand, but this applies equally to progressivism. A meme is the fundamental unit of transmissible culture - a catchy tune, laying napkins on the lunch table, the concept of punishment for wrongdoers. Memes are hosted in sentient minds. A baby's first memes come from its parents. Subsequently, from family, babysitters, toys, telly, other children, books. Memes are virulent to different degrees. An earworm is very infectious, but greeting people with "howdie doodie" is not particularly so. A meme may fade away from a host in time, or be displaced by another meme, or last a long time in the host's mind. Individuals are more or less susceptible to any given meme. This is influenced by the memes they have already. Some memes reinforce each other, and are often found together and transmitted together - for example, the meme for belief in god tends to cohabit with the meme for prayer, and we observe them being transmitted together. The belief in a holy text tends to confer resistance to displacement on proximate memes such as the belief in a personal deity, and also tends to cause the host to expend effort on transmitting the meme complex. These meme complexes are self-sustaining. We might expect meme complexes that include dogma to confer great tenacity on accompanying memes. Good luck convincing a fundie that the earth is old; their memes inoculate them against arguments. So we have the fox news complex: initial infection is through mild right-wing memes, affect the host's behaviour, and increase susceptibility to secondary infections such as "election was stolen" or "9/11 was an inside job". Someone with a very strong base set of memes, such as critical thinking or compassion, is much less susceptible to all the constituent memes of the fox complex. But someone with a weak base meme type - say, consumerism and apathy to books - is more easily infected. I'm not aware of any research in this field. Do please improve my understanding!